Abstract
Science and technology studies made material things participants in social order, and its most influential figure for material stabilisation is Latour's immutable mobile, a form that travels while staying the same. This article argues that STS selected one configuration of material stabilisation and made it stand for the whole, and that the selection follows from the cases with which it began, inscriptions, in which multimateriality withdraws in favour of multisymbolization. Two corrections follow. The first concerns the starting point. Methods that begin with an object already differentiated as the kind of thing it is, including implosion and infrastructural inversion, unfold its relations but cannot reconstruct how that kind of object became possible. Multimaterial stratigraphy supplies that vertical depth, as an accretive history of relations to matter running from secretion and found matter to projective fabrication. The second concerns generalisation. Mobility and mutability, once separated, yield immutable mobiles, mutable mobiles, immutable immobiles and mutable immobiles, and the building emerges as a counterexample in which persistent location combines with controlled change. Further dimensions, among them reproducibility, site dependence, required sameness, repairability, temporal trajectory, radius of remediation and the burden of symbolic stabilisation, interact to generate the institutions that grow around a configuration, as comparisons of the Shard with a tablet, and of bread, banknotes and land, show. Because every configuration remains multimediated, such profiles constrain the problems institutions must address without determining their answers. The immutable mobile is thereby provincialised as one extraordinary solution within a far larger field of made worlds.
I. Introduction: STS Found the Object, But Which Object?
Science and technology studies achieved something that social theory is unlikely to undo. After it, science, technology and social order could no longer be described as though material things were inert background to human action. Maps, laboratories, instruments, microbes, scallops, door-closers and speed bumps, drugs, buildings, standards, documents and infrastructures entered the analysis of society as participants in its ordering (Callon 1986; Latour 1987, 1992; Star 1999). Objects, it became clear, do load-bearing work that face-to-face coordination among living beings could never sustain indefinitely on its own, and a great deal of what looks like social stability turns out to be carried by things (Ecks 2026b).
There is, however, an unnoticed selectivity in the objects that STS made theoretically exemplary. The most influential of them is Latour's immutable mobile (Latour 1986, 1987). A map travels from a ship to a cartographer's table in Europe while remaining stable enough to be combined with other maps. A standardised measurement, a printed table, a specimen drawer and a scientific paper travel in the same way. The immutable mobile is powerful because it joins two properties: it moves, and while moving it stays relatively the same. That conjunction became central because Latour was explaining how centres of calculation could accumulate and act upon distant worlds, and for that purpose it is exactly the right concept.
Why this particular conjunction should furnish a general model of material stabilisation is a different question, and a counterexample makes the difficulty plain. A building may remain on exactly the same site for centuries while being repaired, extended, rewired, replumbed, refaced, repurposed and reclassified. It is, almost perfectly, a mutable immobile. The contrast breaks apart two dimensions that Latour's memorable formulation happened to join: how far something moves, and how much it changes. Once they are separated, immutable mobility becomes one possible combination among several rather than a paradigm.
Mobility and mutability are themselves only two dimensions. A pharmaceutical adds reproducibility, tight standardisation, expiry, property and the remediation of living bodies. A bridge adds site dependence, repair and very long persistence. Food adds perishability and metabolic incorporation. Clothing adds attachment to a moving body and a required deformation in use. Money adds fungibility and symbolically stabilised equivalence. Each domain, once analysed, discloses further ways in which made things differ from one another.
The framework used here treats multimateriality as one of five mediations through which living beings remain coordinated with a world, alongside multisensorial embodiment, being-with, multiversal dwelling and multisymbolization. Multimateriality names the many ways in which living beings recruit, produce, select, displace, alter, combine and eventually anticipate material forms, and its criterion is recursive uptake: a material difference becomes multimaterial only when it enters a living being's further coordination. Multisymbolization, its partner among the poietic mediations, deposits distinctions into durable symbolic form, and symbolic articulations can be stabilised at L4 into portable classifications, categories and inscriptions that travel beyond the coordination that produced them. Every configuration discussed below is a coupling of the two.
The theoretical task that follows is to develop a comparative theory of multimaterial configurations, rather than to replace the immutable mobile with another master object. This article attempts the beginning of such a theory, and it makes two corrections to the dominant orientation of STS. The first concerns where analysis starts: STS methods typically begin with an object already differentiated as the kind of thing it is, and so cannot easily ask how that kind of object became possible. The second concerns what analysis generalises: STS made one consequential configuration of material stabilisation stand for material stabilisation as such. Both corrections retain everything STS has taught about following things through their relations. Both add a question that STS has rarely made primary: what kind of multimaterial achievement is this thing?
II. The First Limitation: Starting With the Object
Joseph Dumit's implosion method is an especially revealing case because it appears maximally expansive. Developed from Donna Haraway's call to implode technoscientific objects such as the gene, the fetus and the chip into the worlds folded inside them (Haraway 1997), implosion asks students and researchers to take an apparently bounded thing and unfold it along many dimensions at once, among them the material, technical, economic, political, historical, bodily, symbolic and educational (Dumit 2014). A pharmaceutical tablet imploded in this way opens onto clinical trials, laboratories, manufacturing plants, intellectual property, prescribers and patients, chemical ingredients, supply chains, regulators and markets. The method is immensely productive, pedagogically and analytically, and it trains a habit of attention that nobody who has practised it loses.
The method nonetheless contains a starting assumption it does not examine: the object has already become available as an object. The investigator begins with the pharmaceutical, the laptop or the cup of coffee, already differentiated from everything around it and already recognisable as the kind of thing it is. Everything subsequent unfolds outward from that given point. A related orientation runs through much of the anthropology of things, from the cultural biography of objects (Kopytoff 1986) to the social lives of medicines, in which an object's career is followed through production, exchange and consumption. Ingold (2007) has objected that studies of materiality begin with objects rather than with the materials and flows from which objects are provisionally formed. His objection is close to the one developed here, but it heads in a different direction, dissolving objects into the continuous flux of materials instead of asking how distinct kinds of relation to matter arose.
Medication shows the limitation decisively. Imploding a tablet cannot by itself explain why some ingested matter became differentiated from food as medicine. It cannot derive the transition from ordinary ingestion to condition-specific material remediation, visible already in chimpanzees who chew the bitter pith of a plant they otherwise avoid when they are heavily parasitised, or in an orangutan who applies chewed leaves to a wound (Huffman 2001; Laumer et al. 2024). Nor can it show the long sequence through which found medicinal matter became selected and displaced matter, processed matter, compounded matter, purified matter and finally projectively fabricated pharmaceutical matter. Each of these relations is present in the tablet, but the tablet does not display them as a sequence, and expanding outward from it, however far, arrives at relations rather than at their genealogy.
The difficulty lies in method rather than in any shortage of historical detail. No amount of expanding outward from an already constituted object necessarily explains the historical constitution of that kind of object. An implosion of a tablet can include the history of the pharmaceutical industry, but it includes that history as one more dimension of the tablet, one more set of relations folded inside it. The question of how medicinal matter as such became distinguishable, and through what accumulation of material relations, belongs to a different axis of inquiry.
Infrastructure studies faces a version of the same limitation, though it came closest to escaping it. Bowker's (1994) method of infrastructural inversion asks the analyst to turn the background of a practice into its foreground, to look at the classifications, standards and pipes that ordinarily go unnoticed. The inversion is powerful because it reveals how much ordinary functioning depends on work that has become invisible. In the terms of the levels of recursivity, infrastructure is coordination that has sunk into L1 through its own success, and Star's (1999) observation that infrastructure becomes visible upon breakdown describes the disruptive route by which such coordination becomes salient again. Inversion is a deliberate opening of it that does not wait for breakdown. Yet it too starts from infrastructures already in place, and it asks what they do and conceal rather than what kind of relation to matter had to become available before anything could be laid down as infrastructure at all.
The contrast can be put simply. Implosion asks what is folded into an object. Multimaterial stratigraphy asks how this kind of object became possible. The two approaches are complementary. Implosion gives horizontal relational density, everything with which this thing is now entangled. Stratigraphy gives vertical depth, both evolutionary and historical: the successive relations to matter that had to become available before anything like this thing could be made, used and recognised. A strong science and technology studies would require both, and at present it possesses mostly the first.
III. Before Technology: The Missing Evolutionary Stratigraphy
STS typically begins extraordinarily late. Its core vocabulary of technology, artefact, tool, infrastructure, instrument and machine already presupposes highly elaborated relations between living beings and matter. Multimateriality begins far deeper. Its history can be reconstructed as an accretive stratigraphy of two preconditions, material dependence and metabolic externalisation, and six partially ordered layers, running from recruited self-produced matter through recruited found, selected and displaced, transformed and combined matter to projective fabrication. Later relations accumulate without abolishing earlier ones, and found and made describe relations between living beings and aspects of material configurations rather than exclusive classes of objects.
The stratigraphy changes the fundamental question STS asks of a made thing. The familiar questions are what a technology does, or through which network it was constituted. A prior question is available: what kind of relation to matter does this configuration instantiate? A mollusc secreting a shell, a hermit crab occupying that shell after the mollusc has died, a person collecting it on a beach, another grinding it into pigment, a jeweller setting it in silver, and an industrial process reproducing its layered microstructure in a synthetic material are not simply different networks involving material actors. They instantiate different multimaterial relations: secretion, recruitment of found matter, selection and displacement, transformation, combination and projective fabrication. The flattened vocabulary of actants cannot easily distinguish these, because it registers participation in a network without registering the kind of relation that makes the participation possible.
Biology has a counterpart of this flattening. Godfrey-Smith (2024) observes that the phrases ecosystem engineer and niche construction are often applied to any organism with significant effects on its environment, whether those effects are the point of what it does or byproducts of feeding and moving, and he argues that describing the history of the Earth requires the distinction between transformation and engineering. Earthworms turn over soil simply by living in it, and a beaver builds a lodge because it wants one. The multimaterial criterion of recursive uptake draws a related though not identical line, since it asks whether a material difference enters the organism's further coordination and not whether it was aimed at. Either way, a vocabulary that treats every effect on matter alike cannot say what a history of material relations needs to say.
This adds a second differentiation to one already made against actor-network theory. Its principle of generalised symmetry obscures categorical differences between nonrecursive, selfrecursive and interrecursive counterparts, between a stone, whose course does not change with how it is engaged, a living organism maintaining itself in response to its own state, and another party whose responses respond to one's own (Ecks 2026a). Recursivity type is in turn independent of recursive relevance, the degree to which a process already intersects the coordination through which people organise their lives: the sun is nonrecursive and immensely relevant, while most microbial species are selfrecursive and relevant to almost nobody (Ecks 2026a). The multimaterial stratigraphy differentiates the material field itself. Even after something has been correctly identified as a nonrecursive material counterpart, the question remains what relation living beings have established with it. A found stone, a knapped stone and a dressed voussoir are equally nonrecursive, and they belong to different layers of multimaterial history.
The stratigraphy also clarifies why STS could begin so late without noticing the cost. The objects it studied most intensively, laboratories, instruments, inscriptions and infrastructures, belong to the uppermost layer, projective fabrication, where multimateriality is recursively coupled with multisymbolization. Those configurations are so saturated with symbolic specification that the older layers beneath them become easy to overlook. Yet the laboratory still depends on found matter, on the transport of samples, on transformations of glass and metal, on combinations of reagents, and on the metabolism of the living beings who work in it. Its instruments belong to a long lineage, since gathering information is among the oldest forms of action and tools have served it for as long as they have served anything else. Godfrey-Smith (2024) notes gorillas probing the depth of a stream with a stick and an otter breathing out bubbles to sniff for traces of chemicals in the air. The instrument and the inscription stand at the far end of that continuum, and they are recognizably descendants of it. Beginning at the top of the stratigraphy is legitimate, provided the top is not mistaken for the whole.
IV. The Immutable Mobile Becomes One Quadrant
The immutable mobile deserves its strongest possible reading. It names a real achievement: the preservation of enough internal organisation, while moving through space, that something can coordinate activity at a distance. Latour (1986) listed the properties that make inscriptions so consequential: they are mobile, immutable, presentable, readable and combinable with one another. Immutability is the source of much of their power, since an inscription that renegotiated its content at each stage of its journey could not let a centre of calculation act on a distant world. An immutable mobile does its work precisely by declining to renegotiate what it carries (Ecks 2026b).
The concept's centrality has a history that can itself be explained. Laboratory studies and the sociology of centres of calculation concentrated on inscriptions: maps, tables, graphs, diagrams and papers. Inscriptions are the multimaterial configuration with the lowest material burden relative to the symbolic content they carry, a thin, light, flat carrier, cheap to reproduce, on which multisymbolization does almost all of the work. Symbolic forms are the most portable forms living beings possess, and an inscription is matter shaped to let them travel. It is no accident that a theory of material stabilisation built from inscriptions arrived at mobility and immutability as its defining pair. The selection of cases produced the generalisation.
The deeper history of writing points the same way. Godfrey-Smith (2024) reports that writing appears to have arisen independently several times, first in Mesopotamia around 3200 BCE, and that its main early uses were recordkeeping and administration and not the fixing of speech. He cites Lévi-Strauss's remark that writing served the integration of large numbers of people into political systems and their grading into classes. Texts also had to become more self-contained before they could serve strangers, since a message that needs its context carried with it cannot travel far. The conjunction of mobility and stability that Latour made central was therefore already what writing was first for.
The two dimensions can be separated, and a simple space results. Relatively immutable and mobile configurations include maps, printed books, coins, standardised components and many pharmaceuticals. Mutable and mobile configurations include food, biological specimens, livestock, donated blood, organs for transplant and unstable reagents, all of which change as they travel and must be moved against the clock of their own transformation, and garments, which must deform continuously as they travel with a moving body. The Zimbabwe bush pump, which de Laet and Mol (2000) showed spreading successfully because its form and even its function adapted to each village, belongs here as well, as a technology that travels well precisely by being fluid. Relatively immutable and immobile configurations include monuments, earthworks, massive foundations and some dams, which stay where they were made and change little over very long periods. Mutable and immobile configurations include buildings, agricultural fields, gardens, roads, factories, sewers and a great many infrastructures, which stay put while being continuously altered.
The space should not be read as four ontological boxes. Mutability depends on timescale, since the most immutable monument weathers, and a coin changes hands daily while its metal wears only over decades. Mobility occurs at several scales, from a pill carried in a pocket to a building whose components are manufactured in one country and assembled in another. STS has itself begun to multiply such possibilities. Law and Mol (2001) distinguished regional, network, fluid and fire-like spatialities, each with its own form of constancy through change. Their topologies remain, however, topologies of how objects hold together in space, and they do not ask about the further dimensions along which made things differ, or about the evolutionary relations to matter that make each configuration possible. The space proposed here is analytical. Its point is that Latour discovered an important diagonal conjunction and, without intending to, made it look like a general principle of material ordering.
The building is the perfect counterexample. Its social usefulness may depend on staying exactly where it is while remaining continuously alterable. A house that travelled while retaining its internal form would lose most of its value as real estate, which is value attached to a location. A house that could never be repaired would soon be uninhabitable. Brand (1994), following the architect Frank Duffy, showed how buildings persist through change by distinguishing layers that change at different rates: the site, which is effectively permanent; the structure, which may last centuries; the skin, replaced every few decades; the services, which wear out faster still; the space plan, which is rearranged; and the stuff, which moves daily. A building learns, in Brand's phrase, by letting its faster layers change while its slower layers hold. Its material achievement is almost the reverse of the immutable mobile: persistent location combined with controlled mutability. It can be named mutable immobility.
Mutable immobility is a form of stabilisation in its own right, and it has generated its own institutions, including surveying, the registration of plots, planning law, building codes and the professions of maintenance and repair, which infrastructure studies has shown to be as constitutive of technological order as invention (Star 1999; Graham and Thrift 2007; Jackson 2014). Gieryn (2002) argued that buildings stabilise social life while remaining open to reinterpretation and reconfiguration, and the conjunction he described is precisely the one the immutable mobile excludes. Science and technology studies, having begun with centres of calculation, took decades to arrive at repair, and it arrived there as a correction to a model it had already generalised.
V. Beyond the Two-by-Two: Dimensions of Multimaterial Difference
The space of mobility and mutability should not be turned into another totalising typology. As soon as another domain is analysed in detail, further dimensions appear, and any list is provisional. At least nine can be identified at present.
Mobility asks how far a configuration can move while remaining itself. A tablet crosses continents; a bridge does not move at all; a tent moves with those who use it, but only between sites where it can be pitched.
Mutability asks how much a configuration can change while remaining the same thing, and it has two sides. A building absorbs extensive change. A banknote tolerates wear but no alteration of its markings. A vaccine that has changed chemically has ceased to be the vaccine. Change is also sometimes required. For a bridge, deformation beyond narrow limits is failure, whereas a shirt must deform almost continuously in use, and much of the history of textiles is a history of making fabrics that deform in controlled ways. Tolerated mutability and required mutability can be specified separately.
Reproducibility asks whether equivalent instances can be generated. The Shard is singular, and a second Shard elsewhere would be a different building. Paracetamol tablets are produced in billions of equivalent units, and their usefulness depends on the equivalence.
Site dependence asks whether functioning depends on one particular location. A bridge is the extreme case, fitted to one gorge, one geology and one river. A tablet is deliberately made to function almost anywhere, once it has been packaged and stored within its specified limits.
Required sameness asks how tightly instances must conform to one another. The stones of a wall need not be identical, and a mason chooses and trims each one for its place. A pharmaceutical dose may be required to fall within tightly specified tolerances, and for biological medicines regulators have had to define a threshold of similarity rather than identity, because exact copying is impossible. Even then sameness is selective: what must stay invariant is a specified set of properties within specified limits, while colour, coating and price may vary.
Repairability asks how much replacement of components can occur while identity persists. A bridge can have its stones replaced one by one and remain the bridge, a working version of the ship of Theseus, and some of the oldest bridges in Europe have been partly rebuilt many times. A damaged tablet is discarded rather than repaired.
Temporal trajectory asks what course of change over time is acceptable or expected. A bridge can weather, darken and acquire historical value as it ages. A medicine is assigned an expiry date after which it ceases, by definition, to be a usable medicine. Food moves along a trajectory from ripeness to rot that storage can slow but not reverse.
The eighth dimension emerged from the analysis of pharmaceuticals: the burden of symbolic stabilisation. It asks how much symbolic specification is required for a material configuration to work as the kind of thing it is. A found shelter requires almost none. A pharmaceutical requires names, strengths, doses, batch numbers, specifications, indications, expiry dates, manufacturing protocols, pharmacopoeial standards and regulatory classifications, and without them a white powder cannot tell anyone what it is.
The ninth emerged from the analysis of clothing: the radius of remediation. It asks how far around itself a configuration modifies the conditions to which living bodies are exposed. A garment modifies the few centimetres of air around the skin, a room modifies metres, a building larger enclosed volumes, and settlements and irrigation systems whole landscapes. A small radius can be carried anywhere the body goes, and a large radius stays behind when the body leaves, so that the radius fixes how a remediation travels and who must keep it in repair.
The dimensions interact, and their interactions are more revealing than any single dimension. High mobility combined with high required sameness generates problems of standardisation, since distributed instances must remain equivalent while passing through different conditions. High reproducibility combined with commercial value generates problems of intellectual property, since whatever can be copied cheaply must be protected by something other than possession. High site dependence combined with long temporal persistence generates maintenance institutions, since a thing that cannot move and must last has to be cared for where it stands. Perishability combined with mobility generates cold chains, which allowed fruit, meat and milk to travel at scales that would otherwise have been impossible and reorganised diet, agriculture and trade in the process (Freidberg 2009). Repairability combined with immobility generates maintenance professions. Required mutability combined with attachment to a moving body generates the problem of fit, which industrial garment production met by converting continuous bodily variation into standardised sizes. A heavy symbolic burden combined with high mobility generates documentation that must travel with the thing: labels, certificates, waybills, passports and serial numbers.
This is more powerful than a static taxonomy because it generates questions about why particular institutions and infrastructures arise. A typology sorts objects into boxes. A set of interacting dimensions predicts the problems that a given configuration will pose to anyone who makes, moves, uses or maintains it, and so predicts, in general terms, the kinds of institutions likely to grow up around it.
Two cautions are needed. The first is that the dimensions are relational. Mobility is mobility for someone, across a distance that matters to someone, and a configuration's profile can change when its uses change: a shipping container is highly mobile in logistics and becomes a site-dependent dwelling unit when it is converted into a house. The second is that the profile of a configuration is distinct from its recursive relevance. Two configurations with identical profiles may matter enormously or not at all to the people around them, depending on how far they intersect coordination that is already at stake (Ecks 2026a). The profile tells what problems a thing will pose if it matters, and relevance tells whether and how much it matters. An adequate analysis needs both.
A short worked case shows the dimensions interacting. A loaf of bread is made from found and cultivated grain, transformed by milling, combined with water, salt and yeast, and transformed again by fermentation and baking. It is mobile over short distances and for short periods, highly mutable, since it stales, dries and moulds within days, reproducible in principle but rarely identical from batch to batch, independent of site once baked, and unrepairable. Its required sameness is loose, so that a slightly different loaf is still bread, and its symbolic burden has traditionally been light. Industrial bread changed the profile. Long-distance distribution demanded slower staling, which demanded new additives and packaging, which demanded ingredient lists and dates, which in turn demanded regulation of what the dates and the lists could claim. Each change in one dimension pulled others after it, and the institutional apparatus of modern food labelling grew from the interaction.
VI. The Pharmaceutical and the Building
A sustained comparison shows the framework at work. The Shard in London and a tablet of a widely prescribed medicine are both strongly projective multimaterial objects. Both existed in symbolic specification, as drawings, calculations, structural formulae and manufacturing protocols, before their complete material realisation. Both require specialised expertise, regulation, capital, labour and heterogeneous materials drawn from many parts of the world. Both belong to the uppermost layer of the stratigraphy. Their multimaterial profiles nonetheless diverge almost point for point.
The Shard is highly immobile and extremely site-dependent: its foundations are fitted to London clay, its form to its plot, its value to its address and its views. It is singular, and its identity does not depend on the existence of other instances. It is continuously repairable, with glazing, lifts, services and interiors replaced over its life while it remains the Shard. It can change substantially while preserving its identity, since offices may become flats and restaurants may close without the building ceasing to be itself. It is expected to persist for decades and perhaps for more than a century. Nothing about it depends on the production of millions of equivalent copies.
The tablet is highly mobile and comparatively site-independent once properly packaged. It is mass-reproduced in enormous numbers of units meant to be interchangeable. It is tightly standardised, with its active content, purity and dissolution specified within narrow limits. It is effectively unrepairable, since a cracked or degraded tablet is discarded. It is temporally bounded by stability and expiry. Its usefulness depends on interchangeability among instances, so that a patient can take any tablet from any pack of the same product and expect the same effect.
Intellectual property becomes intelligible through these profiles. United Kingdom law does protect architectural works, and copyright in them arises automatically, without registration (Copyright, Designs and Patents Act 1988, s. 4). The commercially decisive fact about the Shard nonetheless lies in its matter rather than in its legal status: its material singularity cannot be carried away and sold elsewhere, and possession of the building and its site secures most of what is valuable about it. A pharmaceutical's usefulness depends on its reproducibility, and that same reproducibility makes possession of any particular tablet commercially worthless. The commercially valuable form therefore has to be stabilised symbolically and legally, as a patented molecule, process, formulation or use, a trademark and a regulatory dossier, instead of being secured by material possession.
Pharmaceuticals are no more symbolic than buildings, since the Shard's specifications are as elaborate as any drug's, and its planning permission, building control approvals and title are symbolic stabilisations of great weight. What differs is where symbolic stabilisation has to do its work. For the building, symbols stabilise a singular thing that matter already holds in place. For the tablet, symbols must hold together millions of dispersed instances that matter by itself would let drift apart, and must secure an ownership that matter cannot secure at all. The difference follows in large part from multimaterial configuration.
The two objects also fail differently, and their failures have generated different institutions. When a building fails, through fire, subsidence or structural fault, the failure is local, singular and visible, and it is investigated at the site. When a pharmaceutical fails, through contamination or degradation of a batch, the failure is distributed across thousands of instances in many places, most of them invisible to the people who hold them. Mass reproducibility and mobility together therefore generated the batch number, the lot record and the recall, institutions that allow a distributed failure to be traced back to a common origin and forward to every place the affected instances reached. Nothing comparable is needed for a building, because its failures cannot travel.
A general proposition results. Different material forms generate different requirements for symbolic stabilisation, both in amount and in function. The proposition can be tested against further cases. Money, highly mobile, highly reproducible and valuable only if its instances are accepted as equivalent, generates anti-counterfeiting measures, denominations, mints and central banks. Land, immobile and singular, generates survey, boundary marking and registries of title. Seeds, reproducible by the living things they contain, generated plant variety protection, and technologies designed to prevent farmers from saving seed were patented, though never commercialised. In each case the symbolic and legal apparatus answers to a problem posed by the configuration's profile.
VII. Multimaterial Objects Remain Fully Multimediated
The argument so far could be misread as a new material determinism, in which the physical properties of things dictate the institutions around them. Every multimaterial configuration, however, remains embedded in all five mediations. A bridge is fitted at once to the bodies that cross it, to the coordination among the people who build, use and repair it, to the non-human-made conditions of river, rock and weather, to its own materials, and to the symbolic specifications that govern its making. In a Roman arch bridge, cosmic fit, mesocosmic fit and intersymbolic fit intersect without becoming reducible to one another. Cosmic fit is the bridge's answerability to gravity and the strength of granite, conditions indifferent to every living being. Mesocosmic fit is its answerability to the river, to the bodies that cross it and to the people who build and repair it. Intersymbolic fit is the consistency of its measurements, procedures and rules with one another. Success at one level cannot purchase success at another, and a set of specifications that is perfectly consistent can still describe a bridge that falls.
A pharmaceutical depends in the same way on all five mediations. It depends on embodiment, because it must finally alter a living process in a body that absorbs, metabolises and tolerates it or fails to. It depends on being-with, through the noticing of illness, prescribing, dispensing, caregiving, reminding and trust. It depends on dwelling, on the heat, humidity, light, seasons and distances that no manufacturer controls and against which its stability must be secured. It depends on multimateriality, through its chemical and dosage forms, its packaging, and the clinics, pharmacies, warehouses and vehicles that carry it. And it depends on multisymbolization, through diagnosis, dose, label, regulation and law.
This gives a different account of what Latour called hybrids. Latour (1993) was right that heterogeneous realities are co-present everywhere in the things the moderns claimed to have sorted into nature and society, and wrong to infer that their co-presence requires flattening them into a single register of agency (Ecks 2026a). Heterogeneity can be kept without ontological flattening, since the same object can participate simultaneously in irreducibly different relations, and those relations can be distinguished without being separated. A tablet is chemically active, bodily incorporated, socially transmitted, climatically exposed and symbolically classified, and each of these relations follows its own logic of success and failure. To call the tablet a hybrid registers the co-presence, whereas the five mediations specify what is co-present and how each part can fail.
The multimaterial profile is accordingly one mediation's contribution to the analysis rather than the analysis as a whole. It specifies the problems posed by a configuration's matter, such as its mobility, mutability and required sameness. The answers to those problems are always worked out through the other mediations: through bodies that can or cannot tolerate variation, through relations of trust and authority, through dwelling conditions that test stability, and through symbolic systems that specify, certify and own. Material configuration constrains the problems without dictating the solutions.
VIII. From Networks to Profiles: A Comparative Method for STS
The methodological consequence can be stated directly. STS follows actors, associations, translations, controversies and infrastructures, and none of this needs to be given up. It can be preceded or accompanied by a multimaterial profile of the configuration in question. Such a profile asks a series of questions. Where does the relevant matter originate, and which parts of the configuration are found, secreted, selected and displaced, transformed, combined or projectively fabricated? Which older layers remain active inside the newest ones? How mobile is the configuration, how mutable, how reproducible and how site-dependent? How much sameness is required among its instances, how repairable is it, and what temporal trajectory is expected of it? How far does it remediate the conditions around it, and how much symbolic stabilisation does it require? What happens when it fails, and where does failure typically begin? And then the most important question: which of these properties produced the institutions that surround it?
That last question reverses a familiar STS move. STS has shown, compellingly, how institutions, practices and networks constitute objects. The profile asks, in addition, how different multimaterial configurations generate recurrent institutional problems. Medicines generate quality control, dosage standards, pharmacopoeias, expiry dates, patents and specialised supply chains in part because of the kind of mobile, reproducible and tightly specified remediation they are. Buildings generate planning, surveying, property boundaries, building codes, maintenance and conservation in part because of their site-bound persistence and controlled mutability. Food generates preservation, refrigeration, hygiene regimes and date labelling in part because it is mobile, metabolically incorporated and perishable. Money generates anti-counterfeiting, denomination and convertibility because of yet another profile, one in which fungibility and trust must be maintained across enormous numbers of mobile instances.
The comparative force of the profile shows most clearly where similar institutions arise independently in unconnected places. Land survey and the registration of plots developed in ancient Egypt, where the Nile's floods erased field boundaries every year, in the Roman world and in imperial China. Methods for testing the purity of metal developed in many societies that circulated coinage. Such recurrences are difficult to explain if institutions are traced only through the networks in which each arose, since the networks were largely unconnected. They become intelligible if similar multimaterial profiles pose similar problems, which different societies then solve in historically specific ways.
This is no return to material determinism, since every institutional response remains historically contingent and interrecursive, shaped by the people who contest, negotiate and revise it, and the same problem can be solved in radically different ways or left unsolved. A building's site dependence poses a problem of boundaries, but whether boundaries are held by kin, by a temple, by a state registry or by force is not given by the building. The claim is that material configuration constrains the problems to which institutions must respond, while leaving open how, and whether, they respond.
This is a middle position between technological determinism, which lets technical properties dictate social outcomes, and the tendency within STS to dissolve material constraint into heterogeneous association, in which every constraint is the outcome of a network and none precedes it. The profile identifies constraints that precede any particular network without determining what any network will do with them. It also makes comparison principled. Two objects can be compared by their profiles before either has been followed through its relations, and the comparison generates hypotheses about what following them will find.
A banknote shows what such a profile can disclose. Its substrate is cotton paper or polymer, found and cultivated matter transformed and combined. It is highly mobile, relatively immutable, reproducible in enormous numbers and dependent on strict sameness among instances, since any note that differs noticeably from the others is suspect. It is unrepairable in practice: worn notes are withdrawn and destroyed rather than mended. Its temporal trajectory ends in withdrawal, and its symbolic burden is extreme, because its entire value rests on an equivalence that its matter does not possess. The profile predicts that failure will begin with counterfeiting and with loss of trust, and the institutions surrounding money answer exactly to those problems. One of them deserves notice. Faced with a symbolic burden that matter cannot carry, issuers have pushed part of the burden back into matter itself, through watermarks, security threads, intaglio printing, holograms and transparent windows, material features whose only purpose is to make symbolic sameness difficult to fake. Symbolic and material stabilisation are coupled here in both directions, which is exactly what a profile that attends to all its dimensions at once can bring into view.
IX. What Becomes of the Immutable Mobile?
The immutable mobile should be provincialised rather than discarded. It becomes one especially consequential solution to one multimaterial problem: how coordination can travel while minimising the transformation of what is carried. That problem is enormously important. Science, bureaucracy, empire and capitalism have all depended on its solution, and Latour's analysis of centres of calculation remains one of the most illuminating accounts of how distant worlds came to be governed from a few places.
Other configurations solve other problems. A building solves the problem of how something can remain useful by changing while staying put. A pharmaceutical solves the problem of how millions of copies can remain sufficiently equivalent, and how a substance can travel globally and still act predictably inside highly variable living bodies. A garment solves the problem of how a made environment can move with a body and deform with every movement without losing its hold. A cold chain solves the problem of how something can perish slowly enough to travel. A bridge, repaired stone by stone over centuries, solves the problem of how a material form can retain its identity despite the replacement of its parts. A field under cultivation solves the problem of how a place can be transformed every year and still be the same field.
Once these problems appear, immutable mobility becomes a special case within a far richer field. Its special status in STS can be explained by the cases from which STS generalised, the inscriptions of laboratories and centres of calculation, in which symbolic content travels on the lightest possible material carrier. The immutable mobile is the configuration in which multimateriality most nearly withdraws in favour of multisymbolization. That is why it seemed to capture material stabilisation in general, and why it captures only one extreme of it.
X. Conclusion: From a Sociology of Objects to a Comparative Science of Made Worlds
The achievements of STS are cumulative, and each remains indispensable. STS taught social theory to notice objects. Actor-network theory taught researchers to follow them. Implosion taught them to unfold the relations hidden inside them. Infrastructure studies taught them to notice what disappears into ordinary functioning and what it takes to keep it functioning. Material-semiotic approaches taught them that material and symbolic ordering cannot be separated. The next move builds on all of this by asking a question these approaches have rarely made primary: what kind of multimaterial achievement is this thing?
The article has proposed a double correction. The first concerns the starting point: analysis should reconstruct the evolutionary and historical relations to matter that made an object possible, instead of beginning with the object as though its differentiation were given. The second concerns generalisation: one particularly important configuration, the immutable mobile, should not be treated as though it captured the general logic of material stabilisation. There are immutable mobiles and mutable mobiles, relatively immutable immobiles and mutable immobiles, and even that space is only a beginning, because made configurations differ as well in reproducibility, site dependence, required sameness, repairability, temporal trajectory, radius of remediation and the burden of symbolic stabilisation they impose.
Science and technology studies taught us that things have histories, relations and politics. Multimaterial analysis adds a prior question: what kind of material relation had to become possible before there could be this kind of thing at all? A tablet, a bridge, a building, a book and a banknote occupy different networks, and beyond that they embody different achievements in the long history through which living beings learned to recruit, move, transform, combine, stabilise and project matter. The immutable mobile is one extraordinary achievement within that history, and the history is far larger than any one of its achievements.
Selected References
Bowker, Geoffrey C. 1994. Science on the Run: Information Management and Industrial Geophysics at Schlumberger, 1920–1940. Cambridge, MA: MIT Press.
Brand, Stewart. 1994. How Buildings Learn: What Happens after They're Built. New York: Viking.
Callon, Michel. 1986. "Some Elements of a Sociology of Translation: Domestication of the Scallops and the Fishermen of St Brieuc Bay." In Power, Action and Belief: A New Sociology of Knowledge?, edited by John Law, 196–233. London: Routledge and Kegan Paul.
Copyright, Designs and Patents Act 1988. c. 48. London: HMSO.
de Laet, Marianne, and Annemarie Mol. 2000. "The Zimbabwe Bush Pump: Mechanics of a Fluid Technology." Social Studies of Science 30 (2): 225–263.
Dumit, Joseph. 2014. "Writing the Implosion: Teaching the World One Thing at a Time." Cultural Anthropology 29 (2): 344–362.
Freidberg, Susanne. 2009. Fresh: A Perishable History. Cambridge, MA: Belknap Press of Harvard University Press.
Gieryn, Thomas F. 2002. "What Buildings Do." Theory and Society 31 (1): 35–74.
Godfrey-Smith, Peter. 2024. Living on Earth: Life, Consciousness and the Making of the Natural World. London: William Collins.
Graham, Stephen, and Nigel Thrift. 2007. "Out of Order: Understanding Repair and Maintenance." Theory, Culture and Society 24 (3): 1–25.
Haraway, Donna J. 1997. Modest_Witness@Second_Millennium.FemaleMan©_Meets_OncoMouse™: Feminism and Technoscience. New York: Routledge.
Huffman, Michael A. 2001. "Self-Medicative Behavior in the African Great Apes: An Evolutionary Perspective into the Origins of Human Traditional Medicine." BioScience 51 (8): 651–661.
Ingold, Tim. 2007. "Materials against Materiality." Archaeological Dialogues 14 (1): 1–16.
Jackson, Steven J. 2014. "Rethinking Repair." In Media Technologies: Essays on Communication, Materiality, and Society, edited by Tarleton Gillespie, Pablo J. Boczkowski, and Kirsten A. Foot, 221–239. Cambridge, MA: MIT Press.
Kopytoff, Igor. 1986. "The Cultural Biography of Things: Commoditization as Process." In The Social Life of Things: Commodities in Cultural Perspective, edited by Arjun Appadurai, 64–91. Cambridge: Cambridge University Press.
Latour, Bruno. 1986. "Visualization and Cognition: Thinking with Eyes and Hands." Knowledge and Society 6: 1–40.
Latour, Bruno. 1987. Science in Action: How to Follow Scientists and Engineers through Society. Cambridge, MA: Harvard University Press.
Latour, Bruno. 1992. "Where Are the Missing Masses? The Sociology of a Few Mundane Artifacts." In Shaping Technology/Building Society: Studies in Sociotechnical Change, edited by Wiebe E. Bijker and John Law, 225–258. Cambridge, MA: MIT Press.
Latour, Bruno. 1993. We Have Never Been Modern. Translated by Catherine Porter. Cambridge, MA: Harvard University Press.
Laumer, Isabelle B., Arif Rahman, Tri Rahmaeti, et al. 2024. "Active Self-Treatment of a Facial Wound with a Biologically Active Plant by a Male Sumatran Orangutan." Scientific Reports 14: 8932.
Law, John, and Annemarie Mol. 2001. "Situating Technoscience: An Inquiry into Spatialities." Environment and Planning D: Society and Space 19 (5): 609–621.
Star, Susan Leigh. 1999. "The Ethnography of Infrastructure." American Behavioral Scientist 43 (3): 377–391.